The Gaslighting You Don't See Coming

The 1944 film Gaslight, directed by George Cukor and starring Charles Boyer and Ingrid Bergman, involves a husband who systematically manipulates his wife into believing she is going insane. He dims the gas lights and then denies that they have changed. He hides objects and tells her she has lost them. He isolates her from friends and family while presenting himself as the caring husband managing a fragile wife. By the time the film reaches its climax, she has almost completely lost confidence in her own perception of reality. The film is unsubtle, as a piece of drama. As a psychological portrait, it is more accurate than its melodrama might suggest.

You're not going crazy. But they need you to believe you are, because the alternative is that they're wrong.

From Film to Concept

The term gaslighting entered clinical and popular use gradually over the second half of the twentieth century, but it was Robin Stern's 2007 book The Gaslight Effect that gave it its most accessible clinical framing and brought it into mainstream awareness. Stern, a psychoanalyst and co-founder of the Yale Center for Emotional Intelligence, defined gaslighting as a pattern of interaction in which one person causes another to doubt their own perceptions, memories, and judgments, typically through a combination of denial, trivialisation, and diversion. Her contribution was to shift the focus from the dramatic villain of the film to the quotidian, often unintentional forms of the dynamic that appear in ordinary relationships.

This is the critical reframe. Most popular conversations about gaslighting still invoke the deliberate manipulator, the person who has consciously decided to make you doubt yourself as a strategy of control. This figure exists. But it represents the minority of gaslighting experiences. The far more common pattern involves someone who is not consciously strategising but who is protecting themselves from the implications of your perception of reality. If what you are saying is true, then they are wrong, or harmful, or responsible for something they do not want to be responsible for. Denying your reality is, from their perspective, self-preservation. The outcome for you is the same.

The Three Core Patterns

Stern identifies three primary mechanisms through which gaslighting operates. The first is trivialisation, which involves minimising the emotional significance of your experience. "You're too sensitive." "You're overreacting." "It was just a joke, I can't believe you're making this a thing." The message is not that your perception of events is wrong but that your emotional response to those events is disproportionate, irrational, or a character flaw. This is effective because it exploits the genuine uncertainty most people feel about whether their emotional responses are appropriate. If enough people tell you that your reactions are excessive, you begin to wonder if they are right.

The second mechanism is diversion, which involves changing the subject or redirecting the conversation in ways that prevent the original concern from being examined. "I don't want to talk about this." "You always do this, you always bring up old things." "Why are we talking about me? What about the thing you did last month?" Diversion works because it keeps the conversation moving without ever engaging with the substance of what was raised. The person doing the diverting is often skilled at this in ways that are difficult to name in the moment. The conversation ends with a sense of things being unresolved and a vague feeling that somehow you were the problem.

The third mechanism is blocking, which involves refusing to engage with the reality-testing process itself. "You're imagining things." "That never happened." "I would never say something like that." Blocking is the most direct form of the pattern and, in some ways, the easiest to identify in retrospect. In the moment, it is often destabilising precisely because the confident denial of someone who was present for the same events creates an epistemic problem: one of you is wrong about what happened, and they are very sure it is you.

The Ordinary Relationship Version

What makes everyday gaslighting difficult to identify is that each individual instance of it can be explained away. Any single incident of someone saying "you're overreacting" might be accurate, or at least be a difference of opinion about emotional appropriateness. A single instance of someone changing the subject might be avoidance or social discomfort rather than calculated deflection. A single "that never happened" might be a genuine difference in memory, which is fallible in everyone.

What distinguishes gaslighting from disagreement is pattern and function. Gaslighting is characterised by the consistent direction of the pattern: your perceptions, memories, and emotional responses are consistently the ones that are wrong. Normal conflict involves people taking turns being corrected, with genuine openness on both sides to the possibility that they may have misremembered or misread a situation. Gaslighting involves one person consistently arriving at the conclusion that the other person's reality is unreliable. The asymmetry is the signal.

The function is also diagnostic. Normal disagreement about what happened is oriented toward arriving at an accurate shared account of events. Gaslighting is oriented, consciously or not, toward protecting a particular narrative that serves the gaslighter's interests. The "that never happened" serves to prevent accountability. The "you're too sensitive" serves to prevent the original behaviour from being examined. The diversion serves to run out the clock. When you notice that disagreements always resolve in a particular direction and always seem to serve the same function of protecting one person from scrutiny, that is the pattern worth taking seriously.

Who Is Most Susceptible and Why

Stern makes an observation that is initially counterintuitive: the people most susceptible to gaslighting are often those who are, in other respects, psychologically healthy and emotionally intelligent. Specifically, people who are empathic, who are committed to fairness and accurate self-assessment, and who value maintaining their relationships tend to be disproportionately vulnerable. The reasons for this become clear when you examine what makes gaslighting work as a dynamic.

Gaslighting exploits two psychological tendencies that are, in themselves, adaptive. The first is the genuine epistemic humility that comes from knowing that your perceptions and memories are fallible. Psychologically healthy people know that they sometimes misread situations and sometimes misremember events. This awareness makes them open to correction. The gaslighter exploits this openness by consistently presenting themselves as the more reliable source of truth. The empathic person, wanting to be fair, continues to take this claim seriously longer than the situation warrants.

The second tendency that is exploited is the commitment to maintaining the relationship. People who value their relationships, who care about the other person and want to preserve the connection, will absorb a considerable amount of dissonance before they conclude that the other person is systematically distorting reality. The alternative conclusion, that the person they care about is doing something harmful either deliberately or through a pattern they are blind to, is a conclusion with significant relational consequences. The desire to avoid that conclusion keeps people in the dynamic longer than they would otherwise stay.

This is why the research on domestic abuse dynamics, including work by Lundy Bancroft in Why Does He Do That?, consistently finds that the victims of systematic reality distortion are not, as a group, unusually naive or unusually lacking in self-esteem at the outset. Many begin as confident, competent people. The effect on self-confidence and sense of reality is cumulative and is produced by the dynamic itself, not imported into it.

The Institutional Version

Gaslighting is not confined to intimate relationships. It appears in workplaces, families of origin, and friendships, and some of its most consequential forms are institutional. When an organisation tells an employee that the harassment they experienced did not happen, or was not as serious as they believe, or was a misunderstanding of clearly innocent behaviour, this is institutional gaslighting. When a family systematically denies the emotional reality of a child's experience, reframing abuse as discipline and fear as oversensitivity, this is gaslighting at the developmental level. When a culture tells a particular group that their experience of discrimination is imagined or exaggerated, this is gaslighting at scale.

Jennifer Freyd's work on betrayal trauma is relevant here. Freyd, a psychologist at the University of Oregon, developed the theory that the most psychologically damaging forms of abuse and reality distortion are those perpetrated by people or institutions that the victim is dependent on. The child who is gaslighted by a parent cannot simply leave the relationship. The employee who is gaslighted by an organisation faces significant costs in leaving. The dependent is structurally incentivised to accept the distorted reality rather than to challenge it, because the cost of challenging it includes the loss of the relationship or the institution that provides safety and resources. Betrayal blindness, Freyd's term, is the adaptive psychological mechanism by which people in this position manage to maintain the relationship by not fully processing the betrayal. It is not stupidity or weakness. It is survival.

Distinguishing Gaslighting from Normal Conflict

The term gaslighting has become, in popular usage, somewhat elastic, and this is genuinely problematic. It is sometimes used to describe any disagreement in which one person questions another's interpretation of events, or any situation in which someone denies doing something they are accused of. This broadening of the term strips it of its analytical usefulness and makes it harder for people to identify the actual pattern when it is happening to them.

The markers that distinguish gaslighting from ordinary conflict are worth being precise about. Normal conflict involves genuine uncertainty on both sides about what happened and what it meant. Both people take seriously the possibility that they are wrong. The conversation is oriented toward shared understanding rather than toward a predetermined conclusion. Power in the interaction is roughly balanced, in the sense that both people's perceptions are treated as legitimate data. Resolution, when it comes, involves some acknowledgment of each person's experience.

Gaslighting involves none of these features. One person's perception is consistently the one that is wrong. The other person's version is consistently the authoritative one. Attempts to revisit the conversation or examine the evidence are met with diversion or blocking. The dynamic is characterised by one person's growing uncertainty and the other's consistent confidence. The resolution, when it comes, involves the targeted person accepting the gaslighter's account. If you look back on your conflicts with a particular person and notice that this is the consistent pattern, that is the relevant data.

The Specific Damage to Self-Knowledge

The particular harm of sustained gaslighting is not primarily emotional distress, though that is real and significant. It is the damage to the person's ability to trust their own perception of reality as a reliable guide to action. This is a more fundamental harm than it might initially appear. Our capacity to navigate our lives depends entirely on the reliability of our perceptions and memories. We make decisions based on what we observe and what we remember. We assess danger and safety based on what we perceive. We know whether a relationship is good or bad based on our experience of it.

A person who has been systematically told that their perceptions are unreliable over an extended period begins to experience a genuine erosion of confidence in this navigational capacity. They second-guess observations that would be obvious to anyone else. They rehearse conversations obsessively to determine whether they are misreading them. They look to the gaslighter for guidance about reality, because they have been trained to treat their own perception as the less reliable source. This is the outcome the dynamic produces, whether or not it was intended.

The recovery from this damage is not simple and it is not quick. It involves rebuilding trust in one's own perceptions through accumulated experience of those perceptions being validated by reliable external sources, through therapy that specifically targets the reality-testing capacity, and through careful attention to the difference between the state of genuine uncertainty that good epistemics produces and the performed uncertainty that gaslighting installs. These are related but distinguishable. You can learn to tell them apart. The first step is knowing that you are not going crazy, and that the alternative to that conclusion does not require you to decide that the other person is a monster. It only requires you to notice, clearly and without drama, what the pattern has actually been.